6

2. Literature Review

2.1 Flexible Polyurethane Foam Chemistry

This section focuses on the basic chemical reactions involved in the formation of flexible

polyurethane foams. Since flexible polyurethane foam production requires a variety of chemicals

and additives, this section will review specific chemicals and their importance in the foaming

process.

2.1.1 General Chemical Reactions

Flexible polyurethane foam chemistry particularly features two reactions – the ‘blow’

reaction and the ‘gelation’ reaction. A delicate balance between the two reactions is required in

order to achieve a foam with a stable open-celled structure and good physical properties. The

commercial success of polyurethane foams can be partially attributed to catalysts which help to

precisely control these two reaction schemes. An imbalance between the two reactions can lead

to foam collapse, serious imperfections, and cells that open prematurely or not at all.

2.1.1.1 Blow Reaction

The first step of the model blow reaction (Figure 2.1) involves the reaction of an

isocyanate group with water to yield a thermally unstable carbamic acid which decomposes to

give an amine functionality, carbon dioxide, and heat. In the second step (Figure 2.2), the newly

formed amine group reacts with another isocyanate group to give a disubstituted urea and

additional heat is generated. The total heat generated from the blow reaction is approximately 47

kcal per mole of water reacted,1 along with the carbon dioxide released in the first step and

R N C O H O H R NH

C O

O H +

R N H2 CO 2 HEA T + +

I s o c y a n a t e W at e r Ca r bam i c A c id

A m i n e C a r bo n Di oxide

Figure 2.1 First Step of the Blow Reaction

7

serves as the principal source for ‘blowing’ the foam mixture, though some auxiliary blowing

agents are also usually utilized. Also, since the typical isocyanates utilized in foam production

are difunctional, the second part of the blow reaction serves as a means to chain extend the

aromatic groups of the typically used isocyanate molecules to form linear hard segments.

However, it should be noted that this reaction scheme can also produce covalent cross-linking

points when molecules with functionality greater than two, such as diethanol amine, are added to

the formulation.1

There are other secondary reactions, involving the formation of biuret and allophanate

linkages which could lead to the formation of covalent cross-linking points. In the formation of

biuret, a hydrogen atom from the disubstituted urea reacts with an isocyanate group to form a

biuret linkage, as shown in Figure 2.3.2 The allophanate forming reaction is discussed in the next

section.

2.1.1.2 Gelation Reaction

The gelation reaction, also sometimes called the polymerization reaction, involves the

reaction of an isocyanate group with an alcohol group to give a urethane linkage as shown in

Figure 2.4. The heat of this reaction is reported to be approximately 24 kcal per mole of urethane

R N C O R' NH2 R N C

O

H

N R'

H+

-

Isocyanate Amine Disubstituted Urea

Figure 2.2 Second Step of the Blow Reaction

R N C O + R N C N R'

H H

O

C O

N C N R'R

O

NH

R

IsocyanateDisubstituted Urea Biuret

Figure 2.3 Formation of a Biuret Linkage

H

8

formed.1 Since polyurethane foams usually utilize polyfunctional reactants (typically

difunctional isocyanates and trifunctional polyols), this reaction leads to the formation of a cross-

linked polymer.

The reaction of a urethane group with an isocyanate group to form an allophanate group

is another possible way to further cross-link the polymer as shown in Figure 2.5. In uncatalyzed

systems this reaction is known to be insignificant.2 Also, this reaction is generally not favorable

under the catalytic conditions used for flexible foam production.

It is important to note that both reaction schemes described above occur simultaneously,

and therefore it is critical to control the relative rates of these reactions in order to obtain a foam

with a stable cellular structure and good physical properties. If the blow reaction takes place too

fast in comparison to the gelation reaction, it would result in the cells opening before there is

sufficient viscosity build-up to provide the foam struts with enough strength to uphold the foam,

leading to the collapse of the foam. On the other hand, if the gelation reaction is faster than the

blow reaction, it may result in a foam with closed cells, which is not desirable. The relative rates

of reaction of the isocyanate component with other foam reactants at 25 °C under uncatalyzed

conditions are provided in Table 2.1. These can serve as a guideline to make appropriate catalyst

adjustments to achieve a suitable balance of the two reaction schemes.

+ R N

H

C

O

O CH2 R'R N C O R' CH2 OH-

Isocyanate Alcohol Urethane

Figure 2.4 The Gelation or Cross-Linking Reaction

R N C O R N

H

C

O

O CH2 R'+

NH

R

O

N CR O CH2 R'

OC

Isocyanate Urethane Allophanate

Figure 2.5 Formation of an Allophanate Linkage

9

Familiarity with the above two reaction schemes is adequate to develop a fundamental

understanding of the solid-state morphology which develops in flexible polyurethane foams. As

discussed in Section 2.1.1.1, the blow reaction not only helps in foam expansion, but also leads

to the generation of urea hard segments. The gelation reaction covalently bonds these urea hard

Active Hydrogen Compound

Typical Structure Relative Reaction Rate (Uncatalyzed at 25 °C)

Primary Aliphatic Amine RNH2 100,000 Secondary Aliphatic Amine R2NH 20,000-50,000 Primary Aromatic Amine ArNH2 200-300

Primary Hydroxyl RCH2OH 100 Water H2O 100

Carboxylic Acid RCOOH 40 Secondary Hydroxyl R2CHOH 30

Urea RNHCONHR 15 Tertiary Hydroxyl R3COH 0.5

Urethane RNHCOOR 0.3 Amide RCONH2 0.1

Table 2.1 Reactivity of Isocyanates with Active Hydrogen Compounds1

segments to soft polyol segments. When the concentration of the hard segments exceeds a

system dependent solubility limit, the hard segments phase separate out and form what are

commonly referred to as ‘urea microdomains’. Due to the asymmetric nature of the isocyanates

Polyurea Hard Domain Polyurea BallPolyurea Hard Domain Polyurea Ball

Figure 2.6 Schematic Representation of the Phase Separation Behavior in Polyurethane Foams1

Urea Microdomain

10

utilized in foam manufacture (discussed in detail in Section 2.1.2.1), these microdomains are not

crystalline but have been suggested to possess ordering of a paracrystalline nature.1 In addition,

at higher water contents (and thus at higher hard segment contents), the urea microdomains are

known to aggregate and form larger urea rich structures commonly termed ‘urea balls’ or ‘urea

aggregates’. These urea balls are regions which are richer in urea as compared to the general

surrounding polyol matrix which also contains dispersed urea microdomains. A schematic

representation of this phase-separated morphology is provided in Figure 2.6, and should be kept

in mind while further reading this review. Further aspects of this phase-separation behavior and

its influence on physical properties of foams will be discussed in section 2.3.

2.1.2 Basic Foam Components

There are many different components needed to synthesize a flexible foam. The seven

major ones are isocyanate, polyol, water, physical blowing agents, catalyst, surfactants, and

cross-linking agents.1 The desired end properties of the foam dictate the choice of specific

components along with their required quantities. For example, one way to adjust foam modulus

would be by controlling the percentage of hard segments formed from the water-isocyanate

Component Parts by Weight

Polyol 100 Inorganic Fillers 0-150

Water 1.5-7.5 Silicone Copolymer Surfactant 0.5-2.5

Amine Catalyst 0.1-1.0 Tin Catalyst 0.0-0.5

Chain-Extender 0-10 Cross-Linker 0-5

Additive Variable Auxiliary Blowing Agent 0-35

Isocyanate 25-85

Table 2.2 Formulation Basics for Flexible Polyurethane Foams1

reaction.3 In other cases, it might be required to have a foam with more cell openness – this

would be possible by controlling the type and quantity of surfactant used.1 Table 2.2 lists the

11

components which are commonly involved in a formulation and gives a typical range of

quantities for each component utilized. As can be seen from the table, the quantities of all

components listed are based on the amount of polyol utilized in the formulation. For example,

water is typically used in the range of 1.5-7.5 parts per hundred polyol (pphp). However, the

isocyanate added to the formulation is usually reported by an index number. An isocyanate index

of 100 indicates that there is a stoichiometric amount of isocyanate added to react with functional

groups from the polyol, water, and cross-linkers added in the formulation. In the following

subsections, each type of component will be discussed in detail.

2.1.2.1 Isocyanates

The two most common sources of isocyanate functionalities in foam production come

from toluene diisocyanate (TDI) and diphenylmethane diisocyanate (MDI), of which the former

is more commonly used in North America, where as the latter one has a greater markert demand

in European countries.4 TDI exists in two isomeric forms, as shown in Figure 2.7, both of which

are used in foam production. The two isomers differ mainly in two ways. Firstly, as indicated in

Figure 2.7, the relative reaction rates of the different isocyanate groups on each molecule differ

considerably.5,6 The reactivity of the ortho position in the 2,4 isomer is approximately 12% of

the reactivity of the para position due to the steric hindrance caused by the methyl group.

However, when the reaction temperature approaches 100 °C, steric hindrance effects are

overcome and both the positions react at nearly the same rate. In comparison, the NCO groups on

2,6 TDI have equal reactivities though the reactivity of the second isocyanate group drops by a

factor of around 3 after the first group reacts. The second way in which the two isomers differ is

that the 2,6 isomer is symmetric as compared to the 2,4 isomer and therefore is expected to form

Figure 2.7 Isomers of Toluene Diisocyanate

CH3

NCO

NCO

CH3

NCOOCN(12)

(100)

(56) (56)*

*drops to 17 afterother group reacts

2,4 TDI 2,6 TDI

12

hard segments with better packing characteristics. Chapter 7 in this dissertation addresses the

structure-property relationships of slabstock foams with varied TDI isomer ratios.

The production of TDI, as shown in Figure 2.8, involves the nitration of toluene

followed by reduction and phosgenation steps.1 Routes not utilizing phosgene are commercially

unattractive. Depending on the pathway chosen between these two steps, three industrially

common mixtures of the two isomers of TDI can be generated – 65:35 2,4/2,6 TDI (TDI-65),

80:20 2,4/2,6 TDI (TDI-80), and 100% 2,4 TDI (TDI-100). Of these mixtures, the 80:20 blend is,

by volume, the most important.1 Foam properties can be modified to a certain extent by

modifying the isocyanate used. For example, after suitable catalyst adjustments are made to

enhance the relatively low reactivity, the 65:35 isomer blend has been noted to form foams with

higher load-bearing properties.1

Several issues with regard to the isocyanate have been addressed and are available in the

literature. Knaub et. al. have presented the challenges which MDI based foams face over

conventional TDI foams.4 Dounis et. al. have discussed the effect of TDI index on the

morphology and physical properties of flexible slabstock polyurethane foams.7 Determination of

residual isocyanate in flexible foams via FTIR has also been described.8

2.1.2.2 Polyols

The soft phase of polyurethane foams is usually a polyfunctional alcohol or polyol phase

which on reacting with isocyanate groups covalently bonds with urea hard segments through

para-Nitrotoluene

TOLUENE

nitrate

mixture of mononitro- toluene isomers

nitrate

nitrate nitrate

crystallization crystallization

ortho-Nitrotoluene

65% 2,4-Dinitrotoluene35% 2,6-Dinitrotoluene

2,4-Dinitrotoluene

80% 2,4-Dinitrotoluene20% 2,6-Dinitrotoluene

reduce

phosgenate

reduce

phosgenatereduce

phosgenate

TOLUENE DIISOCYANATE 65/35 Isomer Mixture

TOLUENE DIISOCYANATE 80/20 Isomer Mixture

TOLUENE 2,4 DIISOCYANATE

Figure 2.8 Routes for the Production of Commercial TDI Blends1

13

urethane linkages. Glycols such as ethylene glycol, 1,4-butanediol, and 1,6-hexanediol are

relatively much lower in molecular weight as compared to the polyols used in flexible foam

production. These are more commonly used for chain extension to form hard segments (in

polyurethane elastomers) and therefore will be referred to as ‘chain extenders’. Polyols used for

flexible foam formulations are higher molecular weight (ca. 3000 to 6000 g/mol) and have

average functionalities in the range of 2.5 – 3.1 Polymerization processes allow production of a

wide range of polyols, differing in molecular weight, functionality, reactivity, and chain

structure.2,9 Selecting the right polyol is an important issue, and the choice is governed by the

desired foam properties and economics.

The first polyether polyol which was sold for the production of flexible polyurethane

foams was polyoxytetramethylene glycol.2 Although the use of this polyether polyol resulted in

good overall foam properties, extensive use of the same was restricted due to the high costs

involved. At present, there are two kinds of polyols commercially available for flexible foam

production, hydroxyl terminated polyethers and hydroxyl terminated polyesters. The polyether

polyols are produced by ring opening propoxylation or ethoxylation onto a variety of starting

materials called initiators. Around ninety percent of the flexible polyurethane foam market

utilizes polyether polyols based on propylene oxide in comparison to polyester polyols, because

of their lower cost, better hydrolysis resistance, and greater ease in handling.1 Also, polyurethane

foams, due to their low density cellular structure, expose a large surface area to the atmosphere.

This further makes polyether polyols advantageous over polyester polyols due to the known

greater hydrolytic stability of the polyether backbone. Finally, polyether based flexible foams

contribute lower Tg values, are softer and more resilient, making them suitable candidates for

bedding and seating applications.1

The common polyether polyols used in flexible foam production utilize ethylene oxide

(EO) and propylene oxide (PO) as the repeat units (Figure 2.9). The polyols produced are

typically random heterofed copolymers of EO and PO, though in some cases where high

C H 2 CH2 O CH2 C H O C H 3

x x

Figure 2.9 Repeat Units of the Common Polyether Polyols Used in Flexible Foam Production1

oxyethylene oxypropylene

14

reactivity of the polyol is required, the polyol is EO end-capped. This is because primary

hydroxyl groups are approximately three times more reactive towards isocyanates as compared

to secondary hydroxyl groups.1 The reason behind producing polyols utilizing both EO and PO

monomers is argued as follows. Though polyols based solely on PO have relatively low

reactivities, they are superior as compared to all-EO based polyols in terms of possessing lower

water absorption. On the other hand, EO based polyols become important where water solubility

is required. Thus by making polyols incorporating both repeat units, the resultant polyol gives a

balance of required properties, i.e., lower water swelling is obtained due to the PO repeat units in

the backbone, where as the EO repeat units provide good mixing of the water, isocyanate, and

the polyol. In addition, if end-capped with the primary EO groups, the polyol has a high

reactivity which is of importance for production of high resiliency (HR) foams.

The anionic polymerization of PO and EO for the production of a polyether polyol

involves the successive reaction of an organic oxide with an initiator compound containing

active hydrogen atoms (Figure 2.10).1,10 This requires the addition of the alkylene oxide through

anionic (basic) catalysis or cationic (acidic) to the initiator molecule. Commercial production is

usually carried using a base such as KOH which catalyses the ring opening and oxide addition

CH2

CH

OH

OH

CH2 OH

H2C CH CH3

O

3n+-

KOHCH O

CH2 O

CH2 O

CH2 CH OH

CH3

CH2 CH O

CH3

CH2 CH OH

CH3

CH2 CH O

CH3

CH2 CH OH

CH3

CH2 CH O

CH3

n-1

n-1

n-1

Glycerine Propylene Oxide A Triol

Figure 2.10 Base Catalyzed Production of Poly(propylene oxide)1

CH2

CH2 OH

OHKOH

CH2

CH2 OH

O KH2O+ +

CH2

CH2 OH

O

O

H2C CH CH3+ CH

CH3

CH2 O CH2 CH2OHO

Figure 2.11 Mechanism of Base Catalyzed Ring-Opening Polymerization2

A Tri-functional polyol

15

which is continued until a required molecular weight is achieved. The number of active hydrogen

atoms on the initiator plays an important role in determining the functionality of the polyol, as

can be seen in Figure 2.10. A wide variety of initiators are utilized commercially, such as

ethylene glycol and 1,2-propylene glycol in the production of diols, and glycerine and

trimethylolpropane for the production of triols. It is also common to blend the above mentioned

initiators to achieve desired control over average functionality.1 Some other initiators which

contain a larger number of active hydrogen atoms such as sucrose and sorbitol are also utilized,

generally to increase the average functionality. The mechanism for the polymerization process is

shown in Figure 2.11. The reaction is carried out at ca. 100 °C.2 Water removal is an important

step. Since the epoxide monomers and polyether polyols are easily oxidized, air is excluded from

the manufacturing process.11 When the polymerization is complete, antioxidants are added to

prevent the oxidation of the polyether. The probability of nucleophilic attack taking place at the

first carbon atom of propylene oxide is ten times greater as compared to the attack taking place at

the second carbon atom.2 This leads to the polyether backbone containing predominantly head to

tail units, though some head to head and tail to tail defects are present.

When the polyol production is carried out by feeding a mixture of ethylene oxide and

propylene oxide, a polyol results with random EO and PO units along its backbone, and is

commonly referred to as a ‘heterofed polyol’. In another production scheme, the EO and PO are

fed in a batch wise manner, and this results in a ‘block polyol’.

The production of a pure polyol from a selected initiator is hindered due to the

occurrence of two side reactions. One side reaction involves the formation of a diol on addition

of an oxide to water, which is sometimes present as an impurity in the catalyst, initiator, or oxide

feeds. Thus, a good control over moisture content in the incoming feeds is required to achieve

polyols with desired functionalities and molecular weights. The other side reaction (Figure 2.12)

occurs due to the isomerization of propylene oxide to form an allyl alcohol, which leads to the

formation of monohydroxy molecules with unsaturated end groups, also called as ‘monols’.12

Thus, it can be visualized, that the presence of monols and diols from the two side

reactions mentioned above would play an important role in determining the average functionality

of the polyol, which for flexible foam production is generally desired to be in the range of 2.5 to

3. Various workers have employed different schemes to calculate the average functionality based

on monol and diol content and these can be found in references 1,13. Also, routes to synthesize

16

polyols with lower monol contents and thus better functionality control have been reported in the

literature.14,15

Modified Polyether Polyols

There are three main types of modified polyether polyols each of which is used to make

foams of higher hardness as compared to foams based on unmodified polyols. These are the

polyvinyl-modified polyethers or ‘chain-growth copolymer polyols’, polyols containing polyurea

dispersions or Poly Harnstoff Dispersion (PHD) polyols, and polyols containing polyurethane

dispersions or Poly Isocyanate Poly Addition (PIPA) polyols. Filled polyols are also known to

aid foam processing by improving the cell-opening.

Chain-growth copolymer polyols (CPPs) contain stabilized dispersions of ‘polyvinyl’

fillers which are in-situ “graft” polymerized by a chain-growth mechanism. The preparation of

these involves the break down of an initiator molecule, typically an azobis aliphatic nitrile

compound, to form free-radicals. The radicals then react with the monomer molecules to rapidly

form a high molecular weight polymer. CPPs contain three types of polymer: the vinyl polymer,

unmodified polyether polyol, and some vinyl polymer grafted onto the polyether

macromolecules. Chain transfer agents are often added to prevent the formation of ultra high

molecular weight vinyl polymers. The first commercial CPPs utilized acrylonitrile as the sole

CH2 CH O CH2

CH3

CH O-CH3

xCH3 CH CH2

O+

xCH2 CH O CH2

CH3

CH

CH3

OH + CH2 CH CH2-

O

CH2 CH CH2 O-

CH2 CH CH2 O CH2 CH O CH2

CH3

CH

CH3

OHx

Figure 2.12 Side Reaction Resulting in a Monofunctional Chain (Monol)2

17

monomer. A significant deficiency of using 100% acrylonitrile as the filler was the discoloration

(yellowing) of the foams. This led to the development of styrene-acrylonitrile (SAN) azeotrpoic

copolymer as the favored systems for making the CPPs.1

Polyurea modified polyols, or PHD polyols, are composed of the conventional polyether

polyols and dispersed polyurea particles. The polyurea is formed from the reaction of a diamine

(hydrazine) and TDI in a step growth mechanism. Therefore, unlike the chain-growth CPPs, the

molecular weight build up does not occur in a rapid manner in PHD polyols. The polyols

produced in this manner typically contain 20% dispersed polyureas. The dispersed polyureas also

may react with the TDI during foam manufacture, thereby increasing the cross-linking of the

final polyurethane.

PIPA polyols are conceptually similar to PHD polyols except for they contain dispersed

particles of polyurethanes formed by the in-situ reaction of an isocyanate with an alkanolamine,

for example triethanol amine. In addition to the three main copolymer polyols discussed above,

there are also available the epoxy dispersion polyols, polyisocyanurate dispersion polyols, and

melamine dispersion polyols.1

2.1.2.3 Water

Water acts as a chemical blowing agent and reacts with an isocyanate group resulting in a

primary amine and carbon dioxide as discussed previously in section 2.1.1.1. Increasing the

water content influences both the cell structure and the solid-state morphology of the foam.

Higher water contents typically result in foams with lower density due to the increased blow

reaction, as seen in Figure 2.13 (presented in the next sub-section).2 Also, since the hard segment

content is also increased on reacting more water with the isocyanate, this increases the stiffness

of the polymer composing the foam struts. This effect will be discussed in greater detail in

section 2.3. However, in general, it is observed that increasing the water content while

maintaining other variables constant does not drastically affect the load bearing properties of the

foam. For example, the hardness of the foams corresponding to curve A in Figure 2.13 have been

reported to be in the range of 250-350 N as the water content is varied from 2.0 to 5.0 pphp. This

is due to the fact that the two opposing factors discussed above compensate each other, and by

and large, the load bearing properties remain unchanged.

18

2.1.2.4 Physical Blowing Agents

Although the carbon dioxide produced from the water-isocyanate reaction acts as the

principal source to blow the foam, some formulations also employ physical or auxiliary blowing

agents. These are low boiling solvents, inert towards chemical reactions, and they are generally

used to produce softer foams by reducing the foam density.2 As the foaming reactions proceed,

the temperature reaches ca. 130 °C, and this is high enough to vaporize the low boiling solvents

and provide supplementary gas to expand the foam. Addition of a physical blowing agent while

maintaining the water/isocyanate content constant typically results in larger cells and a greater

degree of cell openness, which results in a decreased foam density generally leading to an

increase in foam softness. This observation has been reported in the literature and is reproduced

in this report in Figure 2.13. However, foams can be produced with similar cellular structures

and varying foam softness. This can be done by partially substituting the CO2 produced from the

water-isocyanate reaction with a physical blowing agent – the foam incorporating the physical

blowing agent would be softer due to a comparatively lower hard segment content.

Until the early 1990’s the auxiliary blowing agent primarily used to produce soft low-

density foams was chlorofluorocarbon CFC-11 (CFCl3).16 This blowing agent was, however,

phased out in 1995 due to the environmental concern it caused regarding the depletion of the

ozone layer, especially over Antarctica.17,18 Nevertheless, at that time, there existed a

replacement, hydrochlorofluorocarbon HCFC-141b (CH3CFCl2), which had performance and

Figure 2.13 Slabstock Polyether Foams water/CFC-11/density/hardness relationship. (A) Nil parts CFC-11. Hardness 250-350 N (B) 5 parts CFC-11. Hardness 250-300N (C)

10 parts CFC-11. Hardness 160-200N (D) 15 parts CFC-11. Hardness 130-160N2

19

handling characteristics similar to that of CFC-11, and was reported to have a depletion effect

which was only 1 to 2 % of that of CFC-11. Although HCFC-141b had a much lower ozone

depleting potential as compared to CFC-11, it was seen as a temporary solution until suitable

nondepleters such as hydrofluorocarbons (HFCs) could be developed for flexible polyurethane

foam production. Also, amongst the currently used HCFCs, 141b has the highest ozone depleting

potential, and is currently targeted to be phased out by Dec. 2002, while other HCFCs have until

2010.

The above chain of events, along with other concerns projected by the EPA, has led

companies which produce or utilize blowing agents to find suitable alternative measures.

Technology for using methylene chloride as a blowing agent exists - however, appropriate

adjustments in the catalyst package are required to overcome processing problems.1 Other

alternatives involving the use of acetone19 and liquid carbon dioxide20 have been suggested in the

literature. Blowing agents such as pentane have been tried to replace CFCs although they are less

satisfactory and also raise flammability concerns. Workers have also proposed the use of certain

additives to achieve softer foams by partially disturbing the formation of the precipitating

polyurea.21

2.1.2.5 Catalysts

Since polyurethane foam production relies on two competing reactions, a balance

between them is required to make foams with good open-celled structures and desired physical

properties. While it is true that these reactions may proceed in the absence of catalysts, they

generally proceed at rates too slow to be practical. Suitable catalysts are thus used to carry out

these reactions in a faster, controlled, and balanced manner. This correct balance is required due

to the possibility of foam collapse if the blow reaction proceeds relatively fast. On the other

hand, if the gelation reaction overtakes the blow reaction, foams with closed cells might result

and this might lead to foam shrinkage or ‘pruning’. Catalyzing a polyurethane foam, therefore,

involves choosing a catalyst package in such a way that the gas produced becomes sufficiently

entrapped in the polymer. The reacting polymer, in turn, must have sufficient strength throughout

the foaming process to maintain its structural integrity without collapse, shrinkage, or splitting.

The role of a catalyst in controlling the balance between the two reactions, as discussed

above, is more conveniently represented by workers in terms of its selectivity.22 Since the

20

number of equivalents of water and alcohol present in the reacting mixture is different, yields of

urea and urethane, which are representative of the blow and gelation reaction respectively,

cannot be compared directly, but require to be normalized with respect to their limiting yields.22

Therefore, the selectivity of a catalyst is defined in terms of ‘normalized’ blowing and gelation

rates.

Normalized Blowing Rate = (% urea yield at time t)/(limiting urea yield) (2.1)

Normalized Gelling Rate = (% urethane yield at time t)/(limiting urethane yield) (2.2)

Then the blow to gel selectivity can be defined as:

Blow to Gel Selectivity = (Normalized Blowing Rate)/(Normalized Gelling Rate) (2.3)

Selectivity values greater than 1 are indicative of a strong preference towards blowing, while

selectivities less than 0.4 are suggestive of a strong gelling catalyst.22 Intermediate selectivity

values indicate more balanced catalysts.

Polyurethane foam formulators generally choose catalysts from two major classes of

compounds – tertiary amines and metal salts, primarily of tin.1,2,11 Since catalysts differ both in

activity and selectivity towards the polyurethane foaming reactions, the two kinds are combined

not only to provide the desired balance of ‘blowing’ vs. ‘gelation’, but also to tune these

reactions according to the needs of the production equipment.

In any chemical reaction, there are certain positions on reacting molecules which are

more susceptible to attack by other added co-reactants. These positions are, therefore, more

likely to undergo a given reaction. Catalysts characteristically function at these positions. In the

formation of polyurethane foams, the catalyst forms an activated complex with the reactants thus

making it easier for the isocyanate moieties to chemically react with the active hydrogen

containing compounds.

Tertiary Amine Catalysts

Tertiary amines, by definition, are compounds which contain a nitrogen atom having

three substituent groups and a free pair of electrons. Though these catalysts are generally thought

of as blowing catalysts, they are known to catalyze the gelation reaction as well.1 The catalytic

activity of the amine is a determined by the availability of a free electron pair for complexation.

The catalysis mechanism involves the donation of these electrons by the tertiary nitrogen of the

catalyst to the isocyanate group leading to the formation of an intermediate complex. The

21

availability of the electrons is a function of both, the steric hindrance caused by the substituent

groups, as well as the electron withdrawing or electron releasing nature of the substituent groups.

Groups which tend to withdraw electrons reduce the accessibility of the electrons and thus

reduce the catalytic activity. N,N-Dimethylcyclohexylamine (Figure 2.14) is an example where

the methyl groups have an electron releasing effect resulting in good catalytic activity.23 The

requirements for good catalytic activity are a) nucleophilic enough to attack the carbon of the

isocyanate group, b) ability to form an active hydrogen amine complex, and c) solubility in water

with the ability to form hydrogen bonds with water.

Since electron accessibility is also usually measured by the basicity, the catalytic activity

is found to generally increase as the basicity increases.23 Thus, a plot of pH vs. catalytic activity

usually yields linear behavior. However, exceptions exist, such as triethylenediamine (Figure

2.15) in which the substituent groups on the nitrogen are pulled back, thereby reducing steric

strain, and thus making the electron pair on the nitrogen atom easily accessible.23 Thus, the

NN

catalytic activity exhibited by this compound is greater than expected from its basicity. A

substantial amount of work has been done to study the functioning of catalysts and their

complexation with the different foaming components. However, much of this work has

investigated the reaction mechanism of model compounds in dilute solutions at fixed

temperatures.11 Also, polar solvents are thought to increase the reactivity of the isocyanates by

stabilizing the polarization of the isocyanate group.11

Besides the electronic effects discussed above, the role of a catalyst is also determined by

other physical and chemical properties. For instance, catalysts which have low boiling points,

such as triethylamine, are readily volatized in the exothermic reactions taking place and are thus

NCH3

CH3

Figure 2.14 N,N-Dimethylcyclohexylamine

Figure 2.15 Triethylenediamine

22

lost from the reacting mixture.23 Once they volatilize, it leads to a decrease in the catalytic effect.

Other catalysts, which contain hydroxyl groups, such as dimethylaminoethanol, chemically react

with the growing polymer chain, and thus are no longer able to find their way to a reaction site.23

This again leads to a loss in the catalytic activity. Therefore, it is often helpful to use a mixture of

catalysts, such that the reactions proceed at a reasonable rate at all times during the foaming as

well as curing stages.

Organometallic Catalysts

While the amine catalysts discussed above exert some influence on the isocyanate-

hydroxyl reaction, organometallic salts favor this reaction almost exclusively and are thus called

gelation catalysts. The catalytic activity is explained by three complimentary mechanisms.1 The

first mechanism describes the activation of the polyol into a tin alkoxide which then reacts with

the isocyanate to yield a urethane linkage.1 The urethane linkage further reacts with a polyol –

thereby propagating the polyol and regenerating the catalyst. In the second mechanism, the

isocyanate molecules get activated and are in turn attacked by the polyol to again propagate the

polymer and regenerate the catalyst.1 The third mechanism involves the formation of a tin-amine

complex which accepts a polyol molecule and further activates the complex.1 This complex then

reacts with an isocyanate group to yield a carbamate linkage. Detailed discussions on these

mechanisms can be found in references 1, 2.

Unlike the tertiary amine catalysts, which usually volatilize during the foaming reactions,

tin catalysts can remain in the foam permanently. However, they are also known to undergo

certain chemical changes. Stannous salts are known to oxidize into their stannic form, which

promotes the oxidative degradation of flexible foams.23

Flexible foams which employ auxiliary blowing agents also require appropriate

adjustments in their catalyst package. The blowing agent sometimes hinders the formation of the

activated complex discussed previously, and is also known to act as a heat sink, since it absorbs

some of the heat generated by the foaming reactions. Thus foams utilizing auxiliary blowing

agents, in general, require higher amounts of catalysts.23 Also, some blowing agents in the

presence of flame retardants are known to have an acidity effect which reduces the basic nature

of the amine and hinders its catalytic activity.23

23

Molded foam applications often employ delayed-action catalysts which allow complete

filling of intricate parts of the mold by delaying the gelation reaction and thus maintaining a low

viscosity of the foaming mixture.1 These catalysts show low activity at room temperature, but

become more activated once the reaction exotherm builds. Examples of these catalysts are

tertiary amine salts in solvents such as water or low molecular weight glycol.

2.1.2.6 Surfactants

Flexible polyurethane foam production relies greatly on the performance of non-ionic,

silicone based surfactants which are added to realize a variety of functions. In fact, the largest

commercial application of silicone surfactants is in the area of polyurethane foams with

worldwide production quantities of 30,000 metric tons/year for the polyurethane foam industry.24

Some of the main functions performed are reducing surface tension, emulsifying incompatible

ingredients, promoting bubble nucleation during mixing, stabilization of the cell walls during

foam expansion, and reducing the defoaming effect of any solids added. Of these functions,

perhaps the most important is the stabilization of the cell walls, without which the foam would

behave like a viscous boiling liquid. The processing window for surfactants is generally in the

range of 0.5-2.5 parts per hundred polyol (pphp), and the actual quantity of surfactant added is

Figure 2.16 Effect of Surfactant Concentration on Foam Stability1

24

dependent on the type of surfactant used as well as on the other constituents of the foam

formulation. Below a certain minimum concentration of surfactant, the foam can result in serious

imperfections such as splitting, densification, or collapse. Addition of more than required

quantities of surfactant has its own drawbacks. This usually results in an over-stabilization of the

foam, resulting in closed cells, which result in a decreased airflow through the foam (Figure

2.16). Also, a high number of closed cells in the foam leads to foam shrinkage on cooling, which

is undesirable.

Although the first “defoamers” used in flexible foams were low viscosity siloxane oils

based on poly(dimethylsiloxane), current surfactants employ graft copolymers of polysiloxane-

polyoxyalkylene.1 The siloxane block is known to lower the bulk surface tension where as the

polyoxyalkylene promotes the solubilization of the surfactant into the polyol and aids in

emulsification of the foaming components. Also, the polyoxyalkylene part is adjusted according

to the type of foam application. This is done by varying the percentage and arrangement of

ethylene oxide units and propylene oxide units which make up the polyoxyalkylene. By doing so,

desired levels of affinity of surfactant for water, and solubility of surfactant in the foam polyol,

can be achieved.1 In addition, surfactants can be tailored to meet required needs of foam systems

by altering their chain architecture (topology). AB, ABA, comb-like, or branched structures are

produced and utilized to meet desired requirements.1 The flexibility of siloxane chemistry

facilitates the production of the broad range of surfactant structures mentioned above.

There are four main types of silicone surfactants, the structures of which are shown in

Figure 2.17. These include silicone oil, ABA block copolymers, graft copolymers, and

surfactants possessing a highly branched structure.25 Table 2.3 lists the structural features of the

surfactant for slabstock as well as molded foams. As noted from this table, the molecular weight

of the surfactants used in slabstock foam production is much higher as compared to those used

for molded foams. This is because stability of the liquid foam as it rises is of primary importance

in slabstock systems. Also, in molded foams, the liquid foam is stabilized mainly by the higher

bulk viscosity of the system which results from the use of higher molecular weight polyether

polyols and the addition of cross-linking agents. The polyethers used in slabstock foam

surfactants are fairly high in molecular weight, ca. 1000-4000 g/mol, as compared to molded

foam surfactants which have polyether molecular weights less than 800 g/mol. The polyethers of

slabstock foam surfactants also possess almost an equimolar ratio of EO and PO contents. This is

25

because surfactants which are based on all EO polyethers do not lower the surface tension of the

polyether polyols and are therefore not practical.26 Surfactants based on all PO polyethers are

less hydrophilic and have shown to be inefficient in slabstock foams.26

Me3SiO-(SiMe2O)x-SiMe3 (a) Me2-SiO-(SiMe2O)x-SiMe2-(CH2)3(EO)a(PO)bOR (b) (CH2)3(EO)a(PO)bOR Me3SiO-(SiMeO)x-(SiMe2O)y-SiMe3 (c) (CH2)3(EO)a(PO)bOR SiMe2-O-(SiMe2O)x-R Me-Si-O-(SiMe2O)y-R (d) SiMe2-O-(SiMe2O)z-R

Figure 2.17 Structural Types of Silicone Surfactants a) Silicone Oil or Polydimethylsiloxane

b) ABA Block Copolymer Structure c) Graft Copolymer Structure d) Branched Structure25

The preparation of a flexible polyurethane foam involves ongoing processes at the

interfacial level as well as within the bulk. Surfactants play a crucial role in influencing both

these processes. The interfacial processes come into play as early as when the initial bubbles are

formed in the liquid. These bubbles are not spontaneously nucleated, but rather require

agitation.27 As the CO2 is evolved due to the reaction of the isocyanate with the water, these

bubbles, which were initially on the scale of a few µm, expand to few hundred µm in size. This

expansion leads to an increase in the net surface area and therefore the surface energy of the

O

O

26

foam. To maintain a low overall surface energy, the addition of silicone surfactants reduces the

surface energy per unit area, i.e., the surface tension.

Table 2.3 Molecular Weights and Copolymer Compositions of Silicone Surfactants

Used for Flexible Polyurethane Foams25

Surfactants are thought to stabilize the air-liquid interface as well as the liquid-liquid

interface. Their addition retards processes such as bubble coalescence, liquid drainage, and

diffusion of gas from smaller to larger bubbles (Ostwald ripening). Bubble coalescence is slowed

down by retarding the excessive thinning of films between adjacent bubbles to an extent which

leads to film rupture. Thinning of films can be offset by transport of surfactant and/or bulk

material to the thinned regions, an adequately viscoelastic surface layer, or by a high bulk

viscosity of the system.

At the air-liquid interface, the adsorption of the surfactant changes the mechanical

behavior of the interface, in particular, its surface tension and viscoelasticity. At a molecular

level, the surfactant conformationally orients at the interface, through bond rotation and bending,

thereby reducing the surface energy. This is due to the very low Si-O bond rotation energy of the

siloxane backbone. The surface tension of the polyols, which form the bulk of the foaming

liquid, is typically in the range of 33-40 mN/m,25 and cannot be considerably lowered by adding

hydrocarbon based surfactants. Silicone based surfactants, however, can reduce the surface

energy to much lower values of 21-25 mN/m.27

During the initial stages of foam formation, at the liquid-liquid interface, surfactants have

been shown to promote the interfacial mixing of the water and the polyol.25 At the latter stages,

when urea hard segments are generated, the surfactant has been shown to stabilize the

precipitating urea phase. This has been shown in studies by Rossmy and coworkers who

demonstrated that when a silicone surfactant was present, the urea phase separation does not lead

to foam collapse, as it does when a surfactant is not present.28

In the bulk, surfactants are speculated to micellize in the water-polyol-isocyanate

mixtures. This hypothesis stems from the observation that a plot of surface tension versus

Application Surfactant molecular weight

Weight percent silicone

Polyether molecular weight

Weight percent EO in polyether

Molded 300-1500 30-100 0-800 0-100 Slabstock 20,000-80,000 15-30 1000-4000 35-65

27

surfactant concentration often display break points similar to those noted at the critical micelle

concentration (CMC) for surfactants in aqueous solutions.25 However, to the authors best

knowledge, there is no experimental evidence in the literature – such as any light scattering/

SAXS/SANS studies, which support the formation of micelles in polyurethane foams.

2.1.2.7 Cross-Linking Agents

Cross-linking agents in flexible polyurethane foams are usually low molecular weight

species with hydroxyl and/or amine groups and have functionalities greater than or equal to 3.

An example shown in Figure 2.18 is diethanol amine (abbreviated as DEOA), a commercially

utilized cross-linker, which is commonly used in molded foam applications as it helps in a faster

viscosity build-up and thus in achieving shorter demold times. Also, since molded foam

applications utilize high molecular weight polyols and slightly higher catalyst doses (as

compared to slabstock formulations), using typical foam surfactants leads to an over-stabilization

of the cell walls. Thus, lower potency surfactants are utilized in molded-foams, to counteract this

over-stabilization effect. Since these surfactants are not potent enough to give dimensional

stability to the foam, the addition of cross-linking agents helps achieve foam stability.

Addition of a cross-linking agent generally leads to a reduction in the stiffness of the

foam. This is because the additional covalent linkage resulting from the cross-linking agent

interferes with the phase separation behavior of the foam. A systematic study initiated by Dounis

and Wilkes29 and continued by Kaushiva and Wilkes30 using DEOA as a cross-linking agent

revealed that the hard segment ordering was lost on addition of DEOA, thus leading to the

observed softening of the foam. Thus it needs to be realized, that even though some components

might be added in small concentrations, the role they play in influencing foam properties can be

very significant and needs to be well understood.

2.1.2.8 Other Additives

Various additives are added to flexible polyurethane foam formulations depending on the

required properties and the end use of the foam. Some additives are added for aesthetic reasons

N CH2 CH2 OHCH2CH2HO

H

Figure 2. 18 Structure of Diethanolamine (DEOA)

28

(for example colorants) where as others are added to improve product performance. Since

polyurethane foams have a significant amount of aromatic content, UV stabilizers are added to

retard the yellowing of foams on exposure to light.1 Bacteriostats and flame retardants are also

added in some formulations. Some other additives include the use of non-reactive plasticizers to

reduce viscosity, cell-openers to prevent shrinkage of the foam on cooling, and compatibilizers

to enhance the emulsification of the reactants.1 The use of antistatic agents to minimize the build

up of static electrical charges is important for foams used to package electronic devices. A

detailed discussion of these additives can be found in references 1 and 2.

2.2 The Foaming Process

FTIR has been extensively used to study the sequence of the foaming reactions. In

general, there is agreement amongst different workers that the water-isocyanate reaction takes

place sooner and faster as compared to the polyol-isocyanate reaction.1,31,32 This is supported by

a growing urea carbonyl absorption at ca. 1715 cm-1 early in the reaction which is observed to

shift to ca. 1640 cm-1 once half the foam rise height is reached. Model studies carried out on

diphenyl urea have indicated that the urea carbonyl absorption in a good solvent (DMF) and a

poor solvent (THF) appears at 1715 and 1640 cm-1 respectively.33 This suggests that a stage is

reached when the polyurea being formed is no longer soluble in the foaming mixture and phase

separation takes place. Bailey and Critchfield observed that the urea formation takes place

quickly with most of it taking place within the first 5 min of the foaming process.31 The urethane

formation, however was not significant in the first 5-10 min, but was found to increase at a

steady rate for the next 30 min. These results were also confirmed by Rossmy and co-workers

who observed that the ratio of isocyanate to water consumption was 2:1 in the early part of the

foam reaction, indirectly indicating that urethane formation was not significant in earlier stages.34

The same workers also confirmed, using reactive and non-reactive polyols, that the heat

generated by the urethane reaction in the earlier stage was negligible.

McClusky and coworkers used a vibrating rod viscometer as a probe to examine the

rheology of the reacting foam mixture.35 Based on their investigation, the reaction scheme was

divided into three regimes. During the first regime, which began from mixing of the reagents and

continued up to the point of cell opening, it was observed that there was a continuous reduction

in the system viscosity due to the increase in temperature resulting from the exothermic nature of

29

the reactions. In the second regime, a rapid increase in the viscosity was observed, due to the

precipitation of the urea which led to the formation of a hydrogen bonded physically cross-linked

network. The third regime displayed a gradual increase in the system viscosity, due to the

formation of the covalent network in the polymer.

The rigidity of rising foams was measured by Bailey and Critchfield based on the BB-

drop test developed by Rowton.31 This test consists of dropping BB’s from a constant height on

the foam sample at different times during the foaming process. The distance traveled by the BB

after hitting the foam can be related to the integrity of the foam. It was observed that the BB’s

sank through the foam until the precipitation of the urea occurred. The phase separation of the

urea, therefore, was responsible to give the foam its structural integrity. It has also been observed

by Rossmy and coworkers that cell rupture took place just after the urea precipitation. In light of

this observation, it has been suggested that the precipitation of the urea destabilizes the foam mix

and aids in cell opening.

Workers have also used different techniques to try and identify the event of cell-opening.

The simplest method to do so is by visual observation of ‘blow-off’ which is marked by a sudden

cessation of the foam expansion and a release of the gas under pressure. Bailey and Critchfield

identified cell opening by measuring the escaped blowing agent concentration above the foam by

using IR spectroscopy.31 Miller and Schmidt used a porosimeter to measure bulk permeability of

the foam, and identified cell-opening with a sudden increase in foam permeability.36 In a more

recent work, Macosko and Neff used a parallel plate rheometer to study cell opening.37 They

observed that the normal force exerted by the expanding foam mixture on the rheometer plates

was a function of both, the rate of foam expansion, as well as the foam modulus. Their work

suggested that the visually observed blow-off of the foam coincided with a sudden drop in the

normal force which marked the cell opening event.

2.3 Morphology

The physical properties of flexible polyurethane foams are a function of both, the cellular

structure, and the phase separated morphology of the polymer comprising the struts of the foam.1

These two factors are intimately related because both are influenced by the forces exerted during

the expansion and stabilization of the foam. There has been considerable effort to try and

understand how these two factors influence the physical properties of the foam such as load

30

bearing, compressive stress-relaxation, creep, and also how these properties are a function of

varied temperature and humidity conditions.1 While testing polyurethane foams, workers have

often found it difficult to separate the effects of cellular structure and the solid state polymer

morphology on the foam properties. For this reason, some investigators have worked on plaques

based on flexible polyurethane foam formulations with an attempt to deconvolute the effect of

polymer morphology on foam properties.3

2.3.1 Cellular Structure

As stated earlier, the properties of flexible polyurethane foams are a strong function of its

cellular structure. A complete knowledge of the cellular structure of a foam would require the

exact size, shape, and location of each cell.38 Since obtaining this information is difficult, and

impractical, certain approximations are employed. Mean cell diameters and average cell volumes

are often used to characterize cell size, since a distribution in cell size is always noted. Earlier

researchers described the shape of the foam cells similar to that of a pentagonal dodecahedron,

which has twelve five-sided faces. However some four- and six-sided faces are also observed in

real polyurethane foams, and thus the cell geometry might be better approximated using the

fourteen-faced tetrakaidecahedron space filling model.1 Another variable of importance for

flexible foams is the degree of cell openness. This is usually characterized using air-flow

measurements.1,30

Optical microscopy as well as SEM has been used to study the detailed features of

cellular structure.1,3,29,30,39 The SEM images for a typical slabstock foam are reproduced in Figure

2.19. It is observed that the foam has an open-celled structure and very few closed cells. Also, a

geometrical anisotropy in the cell structure, parallel and perpendicular to the blow direction can

be observed. The cells parallel to the rise direction appear circular, where as those perpendicular

appear elliptical with their major axis aligned along the foam rise direction. This structural

anisotropy is known to effect bulk foam properties, such as load bearing. A comparison with the

micrographs of a typical molded foam (Figure 2.20), suggests that molded foams have a

significantly greater number of closed cells. This observation has been confirmed by Dounis et.

al., using air flow measurements, which were observed to be 5 ft3/min and 1 ft3/min for slabstock

and molded foams of comparable composition.40 It was also observed in the same study that the

cell struts in the molded foam were thicker, thus resulting in a somewhat higher density foam.40

31

Finally, there was no geometric anisotropy in cell structure noted in the molded foam, probably

because the molded foam operation involves pressurization of the reactants from all directions

into the mold.

Confocal microscopy has been used by workers to collect two-dimensional images of the

cellular structure at different ‘depths’ of the foam, with an objective to provide a realistic three-

a

b

Blow

Blow

Figure 2.19 Typical SEM micrographs of a conventional slabstock polyurethane foam. a) parallel to the rise direction b) perpendicular to the rise direction3

32

dimensional reconstruction of the foam network.41 The use of NMR microscopy for cellular

structure evaluation has also been reported in the literature.42, 43

2.3.2 Polymer Morphology

As discussed above, the cellular structure observed in flexible polyurethane foams plays

an important role in determining its physical properties. If not greater, of equal importance is the

Blow

Figure 2.20 Typical SEM micrographs of a conventional molded polyurethane foam. a) parallel to the rise direction b) perpendicular to the rise direction3

Blow

b

a

33

morphology of the polymer which comprises the solid portion of the foam. Over the years,

workers have utilized several techniques to investigate this solid state morphology. Until the

early 1980’s IR spectroscopy was the primary characterization technique. The last 20 years,

however, have seen the application of SAXS, WAXS, TEM;1,3,7,29,30,40 and more recently AFM44

and XRM45,46 to gain further insight into the unique morphology of these materials at the

molecular, domain, and in most cases at a superstructure level. In conjunction, thermal

characterization using DSC and DMA has often been found helpful.1,3

2.3.2.1 Urea Microdomain Considerations

As discussed earlier in section 2.2, the isocyanate-water reaction proceeds faster as

compared to the reaction between the isocyanate and the polyol. This leads to the formation of

oligomeric polyurea species which are termed as urea hard segments. When the molecular

weight of these urea hard segments exceeds a system dependent solubility limit, thermodynamic

boundaries are surpassed, resulting in a transition from an initial inhomogeneous disordered state

to an ordered microphase-separated state.1,2,3 Workers have suggested using in-situ SAXS

measurements that the microphase separation transition follows the kinetics associated with

spinodal decomposition.47 The characteristic properties of flexible polyurethane foams depend

much less on the covalent cross-linking points present in the polyol phase and more on the

cohesive strength of the microphase separated urea hard domains, which provide physical cross-

linking points. Hydrogen bonds occur readily between the proton donor NH- groups of the

urethane and urea linkages and their electron donor carbonyl groups.1,2 These hydrogen bonds

strongly influence the cohesive strength of the urea hard domains. Other factors; such as the

symmetry, molecular weight, and molecular weight distribution of the aromatic polyurea

segments also strongly influence the nature of packing of the urea hard segments.

Small angle x-ray scattering (SAXS) has been successfully utilized to prove the existence

of the microphase separation discussed above.1,3 An observed shoulder in the SAXS profiles of

flexible polyurethane foams corresponds to an interdomain spacing (average center-to-center

distance between urea hard domains) of ca. 80-120 Å. Workers have also demonstrated using

MALDI mass spectroscopy that the urea hard segments consist of ca. 4-6 repeat units thus

suggesting that the hard domains are ca. 30-60 Å long.48 The ordering or the packing of the urea

hard segments, has been attributed to the presence of bidentate hydrogen bonding. This was

34

confirmed in a recent study by Kaushiva et. al. on analyzing the FTIR spectra and WAXS

patterns of a polyurea powder and a polyurea powder prepared in the presence of a surfactant.49

In that study it was observed that the polyurea powder without surfactant exhibited many peaks

in the WAXS pattern, as would be expected from a crystalline material, and also showed the

presence of an absorbance at 1640 cm-1, indicating the presence of bidentate hydrogen bonding.

However, on preparing a powder with surfactant in it, workers observed via WAXS that the only

periodicity which remained corresponded to a spacing of ca. 4.7 Å, while the 1640 cm-1

absorbance remained unaffected. This study therefore strongly suggested that the local ordering

of the hard segments within the urea microdomains can be examined via the 1640 cm-1 IR

absorbance and the 4.7 Å WAXS reflection.

Moreland et. al. investigated the viscoelastic properties of flexible polyurethane foams

under varied temperature and humidity conditions.50,51 Their work suggested that the observed

increase in creep on increasing the relative humidity was a result of water acting as a hard

domain plasticizer. They also observed that an increase in relative humidity had a greater effect

on the rate of creep at higher temperatures. Interestingly, they also reported that the creep rate

was higher for the higher hard segment containing foams, while maintaining the same initial

deformation level and testing conditions. This difference in creep rates was attributed to the

greater amount of hydrogen bonds available for disruption in the higher hard segment containing

foams. In short, their work elucidated the importance of the presence and stability of urea hard

domains in controlling the properties of flexible foams.

Dounis et. al. investigated the mechano-sorptive behavior of flexible polyurethane foams

undergoing a creep experiment.52 On subjecting the foams to cyclic humidity conditions between

10 and 98 %, the workers showed that the compressive strain increased in subsequent steps, with

larger deformations observed during the desorption portion of the humidity cycling. They

suggested that during the dehumidification process, regions of free volume were introduced in

the urea hard domains, promoting chain slippage and increases in strain. Once more, weakening

of hydrogen bonding was shown to have a marked effect on the viscoelastic properties of flexible

foams.

Not only temperature and humidity, but also the addition of certain cross-linking agents

and additives tends to disrupt the physical associations of the urea hard segments. This was

demonstrated by Kaushiva et. al. on observing that addition of diethanol amine (a commercial

35

cross-linking agent utilized in molded foams, commonly abbreviated as DEOA) had a disrupting

effect on the hydrogen bonding within the urea microdomains.30 Their work suggested that

DEOA primarily resides in the urea microdomains and thus reduces the extent of segmental

packing of the urea hard segments. These changes in structure at the microdomain level thus

provided an explanation for the lower rubbery moduli and lower load-bearing properties

exhibited by the foams containing DEOA.

In another study, Moreland et. al. studied the effect of LiCl as an additive in slabstock

foam formulations.53 The presence of LiCl was shown via WAXS to disrupt the packing of the

urea microdomains. It was also shown using SAXS that the LiCl containing foams were

microphase separated and possessed an interdomain spacing similar to the foam without LiCl.

This study therefore suggested that LiCl acted as a localized ‘hard segment plasticizer’ thus

explaining why the foams which incorporated LiCl exhibited faster rates of compressive load

decay and lower moduli. NMR relaxation times, which can map the motion of an entire polymer

molecule, suggested that in the LiCl containing foams, the hard segments restricted the motion of

the soft segments, as compared to the foams not incorporating LiCl.54

2.3.2.2 Urea Aggregate Considerations

The previous section discussed the importance of urea microdomains in determining

foam properties. There is another structure which is known to exist in flexible polyurethane

foams which needs to be addressed. Urea rich regions, sometimes called ‘urea aggregates’ or

‘urea balls’, which are ca. 2000-5000 Å in diameter, have been observed by workers via TEM1,3

and XRM.45,46 The exact composition and size of these urea aggregates would be expected to

vary from one foam formulation to another. A study carried out by Armistead et. al. on slabstock

foams revealed that as the hard segment content was increased, the urea aggregates increased in

both size and number.3 The micrographs from that study are reproduced in Figure 2.21. It has

also been commonly observed that this urea aggregation behavior is not so pronounced in

molded foam formulations,1,44 a point which will be discussed later.

Not only the composition, but also the mechanism of formation of urea aggregates is not

clear. Currently, there are two mechanisms proposed.1 One mechanism suggests that the

formation of the urea aggregates takes place in two steps. According to this two-step mechanism,

36

phase separation takes place in the first step whereby the urea microdomains are formed. The

second step then involves the diffusion of the microdomains to form larger urea rich aggregates.

The second, and more widely accepted mechanism, suggests that the formation of urea

Figure 2. 21 TEM micrographs of slabstock foams varying in water content3

37

aggregates takes place in regions with higher water-isocyanate concentrations. Since the

solubility of the water in the polyol phase is limited, there may form regions with locally higher

water concentrations. Reaction of the water with the isocyanate in these regions would then lead

to the formation of aggregates which are high in urea content.

Rossmy et. al. have suggested that the presence of these large urea macrophases aids in

the cell-opening of flexible foams.55 They observed that a visible macrophase separation, marked

by a loss in optical clarity, occurred in the bulk liquid just before cell-opening. However, it is not

clear how these urea macrophases (aggregates) induce cell opening. One hypothesis suggests that

some of the urea aggregates may reach a size where they themselves rupture cell windows or

destabilize their surfaces.

The above discussion therefore partially explains why molded foams have more closed

cell windows as compared to comparable slabstock formulations. The high reactivity ingredients

used in molded foams; along with the high molecular weight polyols, which also have higher EO

contents as compared to slabstock polyols; inhibit the formation of large urea aggregates. This

results in the removal of one of the mechanisms which helps in cell opening and thus results in

foams with numerous closed windows. For this reason, it has often been found useful to add

particulate fillers such as copolymer polyols (CPP) to restore this cell opening mechanism.1 In

other cases where these filler particles are not added, it is common industrial practice to

mechanically crush the foam pads between rollers to open the cell windows.

In another study, Moreland et. al. studied the effect of LiCl as an additive on the

morphology and properties of flexible slabstock foams.53 They observed via TEM that addition

of LiCl, even in amounts as low as 0.1 pphp, prevented the formation of urea rich aggregates. In

agreement with the current discussion, it was observed that foams which contained LiCl (i.e. the

foams in which urea aggregation did not take place) had more closed windows as compared to

the foams without LiCl. This observation further emphasizes the destabilizing effect which urea

aggregates might have on cell windows which could lead to their rupture.

2.4 Summary

The reactive processing of water-blown flexible polyurethane foams involves a complex

combination of both physical and chemical events. Attempts to understand the development of a

supramolecular architecture of a solid foam from a liquid mixture of low molecular weight

38

compounds has perplexed researchers working in this area. Based on the findings of workers

over the last five decades or so, the current understanding of the morphological features present

in flexible polyurethane foams is depicted in Figure 2.22. As will become apparent from reading

subsequent chapters, this schematic representation is just a guideline, and might not be truly

representative of the actual morphology of flexible polyurethane foams.

Figure 2.22 Simplified Model for the Morphological Features Found in Flexible Polyurethane Foams1

39

2.5 References

1. Herrington R; and Hock K; Flexible Polyurethane Foams, 2nd Ed., The Dow Chem Co:

(1998) 2. Woods G; Flexible Polyurethane Foams, Chemistry and Technology; Appl Sci Pub Ltd;

Essex, England: (1982) 3. Armistead JP; MS Thesis; VPI & SU: (1985) 4. Knaub PMA; Wiltz EP; Wulay H; J Cell Plast: 33 (1997) 159-184 5. Cooper W; Pearson RW; Darke S; The Industrial Chemist: 3 (1960) 121-126 6. Blender RJ; Handbook of Foamed Plastics; Lake Publishing Corporation; Illinois: (1965) 7. Dounis DV; Wilkes GL; J Appl Polym Sci: 66 (1997) 2395-2408 8. Cole KC; Gheluwe PV; Hebrard MJ; Leroux J; J Appl Polym Sci: 34 (1987) 395-407 9. Hepburn C; Polyurethane Elastomers, 2nd Ed.; Elsevier Appl Sci; London: (1991) 10. McGrath JE; Ring Opening Polymerization, ACS Symposium Series No. 286 (1985) 1-

22 11. Woods, G. The ICI Polyurethanes Book, 2nd ed.; ICI Polyurethanes and John Wiley and

Sons: 1990. 12. Steiner EC; Pelletier RR; Trucks RO; J Amer Chem Soc: 86 (1964) 4678-4686 13. Barksby N; Allen GL; Proc Polyurethane World Cong 1993: (1993) 445-450 14. Schuchardt JL; Harper SD; Proc SPI-32nd Ann Polyurethane Tech/Mark Conf: (1989)

360-364 15. Hinz W; Proc SPI/ISOPA Polyurethanes World Cong: (1991) 519-523 16. Klesper E; Rubber Age, Oct. 1958: (1958) 84-87 17. Decair BR; Pham HT; Richard RG; Shankland IR; Proc SPI-34th Ann Polyurethane

Tech/Mark Conf: (1992) 2-11 18. Dwyer FJ; Knopeck GM; Zwolinski LM; Proc SPI-33rd Ann Polyurethane Tech/Mark

Conf: (1990) 400-406 19. Graff G; Mod Plast: 70 (1993) 32-36 20. Eiben RG; Sulzbach HM; Ferrand JT; Radovich DA; Proc 1995 SPI Polyurethanes Conf:

(1995) 70-73 21. Vandichel JCNE; Appleyard P; Proc SPI-33rd Ann Polyurethane Tech/Mark Conf: (1990)

400-406 22. Listemann ML; Wressell AL; Lassila KR; Klotz HC; Johnson GL; Savoca AC; Proc

Polyurethane World Cong 1993: (1993) 595-608 23. Rusch TE; Raden DS; Plastics Compounding, July/Aug 1980: (1980) 61-74 24. Reed D; Urethanes Technol. 1995: Jan/Feb pp 22-23 25. Snow SA; Stevens RE; Silicone Surfactants, edited by Hill RM; Marcell Dekker Inc. 86

(1999) 137-158 26. Boudreau RJ; Mod Plast: Jan (1967) 133-147 & 234-240 27. Kanner B; Decker TG; J Cell Plast: 5(1) (1969) 32-39 28. Rossmy GR; Kollmeier HJ; Lidy W; Schator H; Wiemann M; J Cell Plast: 17 (1981) 28-

37 29. Dounis DV; Wilkes GL; J Appl Polym Sci: 65 (1997) 525-537 30. Kaushiva BD; Wilkes GL; J Appl Polym Sci: 77 (2000) 202-216 31. Bailey FE Jr; Critchfield FE; J Cell Plast: 17 (1981) 333-339

40

32. Macosko CW; Artavia LD; Low Density Cellular Plastics, Physical Basis of Behavior,

edited by Hilyard NC and Cunningham A; Chapman & Hall, London, UK: (1994) 33-51 33. Hauptmann G; Dorner KH; Hocker H; Pfisterer G; Proc Int Conference, Urethanes

Division; Strassbourg, France: (1980) 635 34. Rossmy GR; Kollmeier HJ; Lidy W; Schator H; Wiemann M; J Cell Plast: 17 (1981)

319-327 35. McClusky JV; O’Neill RE; Priester RD Jr; Ramsey WA; Proc SPI-34th Ann

Polyurethane Tech/Mark Conf: (1992) 535-542 36. Miller TE; Schmidt DL; J Cell Plast: 19 (1983) 326 37. Neff RA; Macosko CW; Rheol Acta: 35 (1996) 656-666 38. Gibson LJ; Ashby MF; Cellular Solids - Structure and Properties, 2nd Ed., Cambridge

University Press: (1997) 39. Rhodes MB; Khaykin B; Langmuir: 2 (1986) 643-649 40. Dounis DV; Wilkes GL; Proc Polyurethanes 1995: (1995) 353-361 41. Hamza R; Zhang X; Macosko CW; Stevens R; Listemann ML; Polym Prepr: 37 (1996)

803-804 42. Chauvaux B; Dereppe JM; Huis R; Magnetic Resonance Imaging; 14 (1996) 937-939 43. Szayna M; Zedler L; Voelkel R; Angew Chem Int Ed: 38 (1999) 2551-2553 44. Kaushiva BD; PhD Dissertation; VPI & SU: (2000) 45. Ade H; Smith AP; Cameron S; Cieslinski R; Mitchell G; Hsiao B; Rightor E; Polymer:

36 (1995) 1843 46. Rightor EG, Urquhart SG, Hitchcock AP, Ade H, Smith AP, Mitchell GE, Priester RD,

Aneja A, Appel G, Wilkes GL, Lidy WE. Macromolecules (in press) 47. Elwell MJ; Ryan AJ; Grunbauer HJM; Lieshout HCV; Macromolecules: 29 (1996) 2960-

2968 48. Yontz DZ; Hsu SL; Macromolecules: 33 (2000) 8415-8420 49. Kaushiva BD; McCartney SR; Rossmy GR; Wilkes GL; Polymer: 41 (2000) 285-310 50. Moreland JC; Wilkes GL; Turner RB; J Appl Polym Sci: 52 (1994) 549-568 51. Moreland JC; Wilkes GL; Turner RB; J Appl Polym Sci: 52 (1994) 569-576 52. Dounis DV; Moreland JC; Wilkes GL; Dillard DA; Turner RB; J Appl Polym Sci: 50

(1993) 293-301 53. Moreland JC; Wilkes GL; Turner RB; Rightor EG; J Appl Polym Sci: 52 (1994) 1459-

1476 54. Moreland JC; Wilkes GL; Moreland CG; Sankar SS; Stejskal EO; Turner RB; J Appl

Polym Sci: 52 (1994) 1175-1180 55. Rossmy GR; Kollmeier HJ; Lidy W; Schator H; Wiemann M; J Cell Plast: 13 (1977) 26-

35